6025 Background: Patients (pts) are referred to high-volume (Hi Vol) hospitals (hosps) for high-risk cancer surgery (pancreatectomy, esophagectomy) because peri-operative mortality is significantly lower at such hosps. For moderate-risk procedures (mortality 1%–5%), the benefit of Hi Vol hosps is significant, but modest. The value of referral for these common procedures is unclear, but alternatives are lacking. We hypothesized that, independent of procedure volume, hosps with high staff-to-bed ratios would have lower mortality after moderate-risk cancer surgery. Methods: We studied inpatient claims of pts who underwent 3 moderate risk procedures (segmental or wedge resection of lung, 646 lung ca pts; partial or total cystectomy, 1510 bladder ca pts; partial colectomy, 14,948 colon ca pts) between 1/99 and 12/01 in all Texas hosps, except federal. We studied the relationship between inpatient mortality and hosp staffing, using a random effects logit model to account for clustering of pts within hosps. Staffing factors included registered nurse (RN)-, total nursing staff-, and respiratory therapist-to-occupied bed ratios. We adjusted for confounding pt factors (age > 75 yrs, African-American race, Hispanic ethnicity, low income, male sex, chronic comorbidity, metastases, cancer site) and hosp factors (teaching status, rural location, for-profit ownership, total annual surgeries). Results: 17,104 pts were treated at 273 hosps. 51% were male, 37% were >75 yrs, 11% were African-American, and 13%, Hispanic. Inpatient mortality was 4% for segmental lung resection, 2.2% for cystectomy, and 3.2% for partial colectomy. After adjusting for confounding, treatment at a Hi Vol hosp (OR=0.74; p = 0.001) and a high RN-to-bed ratio (OR=0.80; p = 0.02) predicted significantly lower inpatient mortality. These findings were consistent overall and for each procedure individually. Total nurse and respiratory therapist staffing did not predict mortality. Conclusions: A high RN-to-bed ratio is a significant predictor of low peri-operative mortality after moderate risk procedures. Increasing RN staffing could improve outcomes in low vol centers. No significant financial relationships to disclose.